Booms, Recessions, and the Economically Disadvantaged

William Poole*
President, Federal Reserve Bank of St. Louis

School of Business & Management
University of Arkansas at Pine Bluff
Pine Bluff, Ark.
March 5, 2003

*I appreciate assistance and comments provided by my colleagues, especially Howard J. Wall, Research Officer, at the Federal Reserve Bank of St. Louis. I take full responsibility for errors. The views expressed are mine and do not necessarily reflect official positions of the Federal Reserve System.

Booms, Recessions, and the Economically Disadvantaged

It is a great pleasure to be here today, especially given that I've gotten to know Chancellor Davis through his service on the board of the Little Rock Branch of the Federal Reserve Bank of St. Louis. The Federal Reserve relies on the dedicated service of people like Dr. Davis who are members of Federal Reserve boards all over the country. These board members are part of a network of highly informed citizens who provide information to us about business conditions, and who help us to communicate better with communities large and small.

My focus today is the labor market, and especially the situation faced by those who are economically disadvantaged. People on the bottom rungs of our society always have the greatest problems at a time like today, when economic growth is low. The U.S. economy entered a recession in March 2001, and unemployment began to rise. The National Bureau of Economic Research, which officially dates business cycle turning points, has not yet declared an end to the recession. However, the economy has been growing slowly for over a year; I'll hazard a guess that the recession ended in late 2001 or early 2002. That said, growth has not been robust and consequently the number of jobs has been about flat after declines in 2001. Unfortunately, disadvantaged members of our society are typically disproportionately affected by recessions, including the recession of 2001. I'll document some of the facts of that sad regularity.

Although the recession has been painful for many, I do not want to concentrate entirely on that aspect of the situation in the labor market. As I will explain, reviewing long-run labor market trends suggests two main findings. First, disadvantaged workers have in fact made great strides over the last quarter century. Second, improvements in education and earnings have been particularly marked for black women.

Before proceeding, I want to emphasize that the views I express here are mine and do not necessarily reflect official positions of the Federal Reserve System. I thank my colleagues at the Federal Reserve Bank of St. Louis, especially Howard Wall, for their assistance and comments, but I retain full responsibility for errors.

Long Term Trends

Recent decades have seen significant improvements in the relative well being of historically disadvantaged groups, particularly women and blacks. By the broadest measure of well being—income—these groups have made advances in both absolute and relative terms. I'll focus on real median income—defined as income adjusted for inflation of those employed and aged 15 and older. Between 1970 and 2001, the real median income of black men rose by 27 percent, which was nearly three times the growth for white men. Over the same period, real median income for all women rose by 60 percent and for black women by 70 percent. Average income growth for white and black women was, respectively, six and seven times the growth seen by the median white man.

As a consequence, the very large income gaps that existed in 1970 have fallen somewhat. The average woman in 1970 had a yearly income that was only 34 percent that of a man, but by 2001 this gap had been reduced by two-thirds. For black women, the income gap has been more along gender, rather than racial, lines. In 1970, the average black woman earned 91 percent what a white woman earned in a year. This relatively small gap has, for the most part, disappeared. By 2001, the average income for black women was 98 percent that of white women.

For black men, the rise in relative income over the last thirty years or so was not nearly as dramatic as it was for women. In 1970, the median yearly income of black men was 59 percent what it was for white men. By 2001, it had become 71 percent of the average white male's yearly income. Although this indicates some progress, there is obviously a long way to go.

Another way to look at the gains that have been made by disadvantaged groups is to look at those at the bottom of the income scale. A society's success cannot be measured solely by how it provides for the average person, but also by how it provides for those at the low end. According to the changes in the poverty rate, the story for blacks is much the same as I have been outlining with regards to income. In 1970, the poverty rate for whites was 9.9 percent, while for blacks it was more then three times as high—33.5 percent. In 2002, the poverty rate for whites differed little from its 1970 level, but for blacks it had been cut by one third. Again, while the data show that substantial progress has been made, they also show how much more needs to be done.

So far, I have been speaking only about the progress that has been made in terms of broad measures of economic well being. But to truly understand these gains, we must understand what progress has been made in providing the means of achieving them. In particular, if we look at the dramatic improvements in educational attainment, we can see the source of much of the progress, along with the distance we still have to go.

One major achievement has been a narrowing of the racial gap in the attainment of a high school diploma. In what follows, I'll concentrate on educational attainment for those 25 and older. In 1970, 55 percent of whites, but only 34 percent of their black counterparts, had completed high school. By 2000, the gap had narrowed dramatically, so that 79 percent of blacks, compared with 84 percent of whites, had completed high school.

Because a college education increasingly has become the key to economic success for individuals, it is also important to note how access to higher education has contributed to the income trends. Here, the racial gap has not closed nearly as much as it did for high school attainment but, then again, there was much more to be done at the outset. In 1970, a black male was only about 30 percent as likely to have a college degree as a white male. By 2000, black males were nearly 60 percent as likely to have a college degree.

For women, the more significant changes in educational attainment have occurred with respect to higher education, rather than to secondary education. Even thirty years ago there really wasn't much of a gap between the sexes in the rates at which they completed high school. But at the college level, the story has been quite different, and more complicated. In 1970, a white woman was only 57 percent as likely as to have a college degree as a white male, but by 2000 this statistic had increased to 84 percent. For black women, relative to black men, there really has not been a gender gap. Even before 1970, a black woman was about as likely as a black man to have a college degree. In fact, for the last few years, the percentage of black women who have college degrees has been consistently higher than the percentage of black men who did.

The relatively greater educational attainment and income of black women compared to black men is a recurring part of the story: Over the last thirty years or more, the gains in economic opportunity and well-being have been relatively greater for black women than for black men. While, on average, black men have seen progress, black women have seen more. In addition to educational attainment, gains of black women are apparent in employment outcomes. By most measures, black women have succeeded in reducing or eliminating employment gaps between themselves and white women. But for black men, the gaps have been larger and much more persistent.

One way to highlight the complexities of the situation is to look at the shares of the various adult populations that have are employed at different points in time, where adults are defined as those who are 20 and older. The ratio of employment to population—which I'll call the "employment ratio"—is a good indicator of the rate at which members of a group actively participate in the economy. In 1972, the employment ratio for white males was 79 percent, compared with 73 percent for black males. By 2000, the employment ratio for white males had fallen by 5 percentage points while for black males the ratio had fallen by 7 percentage points. Thus, despite the relative gains in educational achievement for black men, the gap between white and black men in the percentage who are employed has actually grown over the last thirty years.

The story for black women has been very different. In 1972, the share of black women who were employed was actually 6 percentage points higher than for white women. Since then, women of all races have been drawn increasingly into the workforce, rising from 41 to 58 percent of the population between 1972 and 2000. Although it wasn't true for the entire period, in 2000 the share of black women who were employed was still higher than the share of white women who were.

The Business Cycle and the Disadvantaged

It is important to note that the progress I have described did not occur smoothly over the period. Between 1970 and 2000 the economy went through some rocky patches. We saw five recessions, a huge runup in inflation in the 1970s and a long fight against inflation in the 1980s and 1990s. Following a post-war high in unemployment in 1982, the economy then grew on a sustained basis, except for a brief recession in 1990-91. We again experienced a mild recession in 2001, although this one has a lingering feel to it given that employment gains were minimal 2002 and remain so to the current day.

This economic turbulence slowed the progress of disadvantaged groups, who traditionally bear disproportionate burdens during economic slowdowns. Although the progress in educational attainment was fairly consistent, the progress in economic well being was actually set back during several periods.

In particular, although white and black women didn't have dramatically different average incomes in 1970 or in 2001, black women actually lost ground relative to white women throughout the 1980s. Although the rise in the income of white women relative white men was fairly continuous and recession proof, the dramatic improvement in black women's income did not begin until 1989. Similarly, for black men, until 1989 there was practically no movement in income relative to white men. Almost the entire gain occurred in the 1990s. Further, the fall in the poverty rate for blacks was almost entirely a 1990s phenomenon, as it did not go below 30 percent until 1995.

Recessions wreak havoc on the progress of disadvantaged groups, particularly blacks. This unfortunate fact can be seen by looking at the effects recessions have had on the gaps in employment ratios. The exact numbers differ from one recession to another, but the clear regularity is that the gap between the employment ratios of black and white men tends to rise during recessions and fall during expansions. But expansions close the gap more slowly than recessions open it. Between 1972 and 2000, for each year of recession it took three years of expansion for the gap to return to its pre-recession level.

Recessions have been even more destructive to the relative progress of black women. In fact, for each year of recession over the last 33 years, it took about four years of expansion for the gap between black and white women's employment ratios to return to its pre-recession level.

Therefore, one of the keys to improving the relative status of disadvantaged groups is for the economy to maintain steady and sustained economic growth. We at the Fed are convinced that the critical contribution we can make toward maximum sustainable economic growth is to maintain low and stable inflation-price stability, for short. When prices are unstable, businesses and households face more uncertainty about the future, making it more difficult for them to plan efficiently. When people plan inefficiently, unavoidable mistakes are more common, which leads to greater variability in growth and employment. Price stability was a necessary ingredient of the 1990s expansion. If inflation hadn't been kept in check throughout the decade, the result would have most certainly been slower growth, and slower progress for disadvantaged groups.

I believe that improved monetary policy since the 1970s has contributed to a reduced frequency and severity of recessions. The business cycle expansions of 1982 to 1990 and 1991 to 2001 were both much longer than the average cycle expansion over the period for which we have a business cycle chronology, which starts in the 1850s. The same statement holds if we confine attention to the period since World War II. Moreover, the recessions of 1990-91 and 2001 were considerably less severe than the average recession. Sustaining business cycle expansions does help to cement the progress of disadvantaged groups, and reducing the severity of recessions reduces the magnitude of the setbacks that occur during recession.

Because sustained expansions as occurred in the 1990s are important to improving the well being of all Americans, we should take a closer look at that decade to see how the benefits of its economic expansion were spread. I'd like to broaden the discussion a bit to include disadvantaged groups other than women and blacks. In particular, I would like to see how some groups that began the 1990s in the worst economic shape, including teenagers and those at the lowest ends of the education and income distributions, also enjoyed substantial gains.

The 1990s Expansion

Because it is the most widely used indicator of labor-market performance, let me start with the drop in the unemployment rate. After peaking at 7.8 percent in June of 1992, the overall unemployment rate fell steadily throughout the 1990s, reaching 4.0 percent by the end of 1999, where it hovered for another year. When we disaggregate these unemployment numbers, it becomes apparent that the expansion was very beneficial for groups that began the period in the relatively worst situations: blacks, teenagers, and the less educated.

The unemployment picture for blacks was pretty grim in 1992, when the unemployment rate for this group averaged 14.2 percent. That rate fell to 7.3 percent by the end of 2000, which was lower than at any time since 1972. Interestingly, black unemployment continued to fall for more than a year after the unemployment rate for whites had leveled off. So, although the unemployment rate for whites remained lower than for blacks, continued economic growth meant that the unemployment gap between whites and blacks kept narrowing. The decline in the black unemployment rate from 14.2 percent to 7.2 percent between 1992 and 2000 is a measure of our nation's progress during the 1990s, but the remaining substantial gap between black and white unemployment is a measure of the distance we still had to go.

The 1990s expansion also meant good news regarding the teenage unemployment rate, defined as the rate for those 16 to 19 years old. The teenage unemployment rate in 1992 averaged 20.1 percent for all races. By the end of the decade, the rate had fallen to around 13 percent, a 30-year low. Black and white teenage unemployment rates were both at their lowest levels in 30 years, although the teenage unemployment picture for whites was still much better than that for blacks.

Well-educated workers, of course, were highly valued by employers in the 1990s. Nevertheless, the less educated clearly reaped benefits from the economic expansion. In 2000, the unemployment rate for those older than 25 who did not have a high school diploma averaged 6.3 percent, having fallen from a high of 12.2 percent in mid-1992. For those with a high school diploma, but no college training, unemployment averaged 3.5 percent in 2000, having fallen from 7.3 percent in mid 1992.

Although these unemployment rates indicate success in bringing those without college degrees into employment, the unemployment rate for college graduates of only 1.7 percent in 2000 shows the tremendous importance of improving the education of citizens. Even today, according to the latest statistics for January of this year which showed a national unemployment rate of 5.7 percent, the unemployment rate for those 25 and older with a college degree was 3.0 percent. The unemployment rate for those with some college was 4.8 percent and for those with less than a high school diploma was 8.5 percent.

Unemployment rates tell only part of the employment story. During any period when employment opportunities are expanding, two things happen: First, more people become employed; and, second, more of those reported as being not in the labor force decide to enter, or reenter, the labor force. Although both of these effects are important, newspaper reporters and TV newscasters tend to look only at the first, and to ignore the second.

As I discussed earlier, employment ratios highlight one of the great successes of the 1990s expansion—bringing increasing shares of women and blacks into employment. Between 1992 and 2000, the employment ratio for white women rose by 3.1 percentage points; that of black men rose by 3.5 percentage points. Compare these numbers to those for the 20 previous years. Between 1973 and 1992, the share of black men who were employed actually fell by 8.7 percentage points. But the really astounding experience belongs to black women: Between 1973 and 1992, the share of black women employed grew by 7.1 percentage points, and then by another 7.7 percentage points between 1992 and 2000.

For another perspective, we can divide all of the households in the United States into three real income categories: those in the Low group have incomes below $25,000, those in the Middle group have incomes between $25,000 and $50,000, and those in the High group have incomes above $50,000. In 1992, 34 percent of households were in the Low group; by 2000, only 29 percent of households were in the Low group. Households in the Middle group fell from 29 percent to 28 percent of all households. The High group, therefore, went from 37 percent to 43 percent of all households. I should note that this evidence is by no means definitive, but it does illustrate that the sustained economic expansion raises the economic well being of many of those at the low end.

2000-2003 Slowdown

So far, I've said little about what has happened since 2001. For one thing, many of the data series that I have referred to are not yet available for the last two years. But, more importantly, I wanted to have a separate discussion of the most recent years because they have been quite different from the years that preceded them. Most notably, the national unemployment rate began to rise in October of 2000—reaching 6 percent by the end of 2002. The economy entered an officially designated recession in March 2001. Even though the economy has been expanding for a year or so, the labor market has not shown the rapid improvements that often accompany recovery periods.

Let's begin by disaggregating the changes in the unemployment rates by race and education level. If past experience is a guide, we would expect to see that the recent slowdown has had larger effects on the less educated and on blacks—and it has. The total unemployment rate for those aged over 25 rose by 1.5 percentage points between the third quarter of 2000 and the third quarter of 2002, while the unemployment rate for those with college degrees rose by 1.1 percentage points. Also, the black unemployment experience has been somewhat worse than this, with the overall unemployment rate rising by 1.8 percentage points and the rate for college graduates rising by 1.6 percentage points.

It is interesting to note, however, that the picture is different for black men compared to black women. For black men, the increase in the unemployment rate for those with college degrees rose very little. For black women, the opposite occurred as the increase in unemployment for those with college degrees actually rose by more than the overall unemployment rate.

Economic slowdowns usually cause people not only to lose their jobs but also to leave the labor force altogether. During this slowdown, women who lost their jobs were more likely than men to leave the labor force rather than look for a new job. Thus, changes in unemployment rates understate the effects of the current slowdown on women and obscure the evidence that this slowdown has been rather different from previous ones.

An important difference is that the employment ratio for women with college degrees fell substantially more than it did for women overall. For white women, the overall employment ratio fell by one-half of a percentage point, whereas for women with college degrees it fell by three times this much. For black women the overall rate fell by 1.5 percentage points, while for those with college degrees it fell by 3.7 percentage points.

This pattern for women was the reverse of what has occurred for men. For white and black men, the employment ratios for those with college degrees rose slightly less than for those without them.

The reasons for this divergence in results for men and women is that the job losses since mid 2000 have occurred primarily in the manufacturing sector, where men without college degrees are predominant, and in the business services sector, where college-educated women are abundant. In fact, these two reasons are actually closely related because the business services industry provides services to the manufacturing sector. In particular, it provides temporary workers, who are more likely to be women because of the time flexibility that it provides, and also are more likely to be among the first to be let go in a slowdown. In a sense, then, the disproportionate burden that the current slowdown has placed on college-educated women is a result of the growth of women's opportunities during the 1990s expansion. Labor markets became more flexible to accommodate the needs of educated female employees during boom times, but this flexibility turns against women during slowdowns.

Summing up

I know that I've given you a huge dose of numbers; even if you had been madly taking notes you could not have gotten them all. If you are interested in looking again at the numbers, you'll be able to find this speech on the St. Louis Fed web site shortly. I'll summarize the main themes.

The 2001 recession has been typical in that the burden imposed on disadvantaged members of society has been disproportionate. But the data also show clearly that education makes a critical difference. Each of you probably understands that fact when you observe the experience of friends who are not pursuing college or university studies.

At the same time we acknowledge the disproportionate impact of a slack labor market, we should also celebrate the tremendous progress the nation has made over the last quarter century. Income and educational disparities by sex and race have been declining. The efforts the nation has made really do show up in the data. The job is far from over, but we have a good start.

Since employment peaked in early 2001, the labor market situation has been difficult for many of our fellow citizens. I suspect that many of you have had first-hand experience with the problem of finding summer jobs, and the difficulty graduating seniors have had in finding suitable positions. The Federal Reserve itself is not immune from these problems, as evidenced by its recent announcement that it will be closing 13 check-processing operations across the country, affecting 1300 jobs.

But in time, I hope soon but do not know for sure, the economy will be expanding again and employment will grow. The U.S. Economy is highly entrepreneurial, and growth its normal state. The Fed will do its best to contribute to this growth.

Thank you, and I'd be pleased to take your questions.

Back to top